The point is that this "benchmark" is impossible to actually see, and the numbers are ridiculous.

No, I also do not believe that the C++ version is 5000 times faster than the Python version without some sort of validation that I can look at.

I have spent most of my life writing C++. I am perfectly willing to believe that a C++ program is 30 times faster than a python program, but not 5000 times.

If the writer had given us anything, anything, anything at all except a screenshot with almost no information I could evaluate the benchmark for reasonability, or even run it myself.